MFA methodology and definitions
Pixalate bases its MFA ratings on traffic it actually observes across its global data pool. It doesn’t rely only on crawling, because crawlers can be gamed. This page explains which signals Pixalate measures, how they become a risk rating, and what each term means.
How detection works
- Pixalate measures a set of signals for each website, mobile app and CTV app. The signals differ by environment (see below).
- For each signal, Pixalate calculates baselines by analyzing quantiles across all ad impressions per website or app in its global traffic pool.
- A property is flagged as likely MFA when its signals fall outside those baselines. The thresholds are set to catch truly aberrant behavior.
- The result is an MFA risk rating of high, medium or low.
Signals by environment
The app signals (age, IVT, popularity score, reviews) are inputs to the rating. The Enrichment API’s mfa widget returns the MFA risk and ad refresh rate for apps, and the MFA risk plus all four website signals for domains. See the domain, mobile app and CTV app fields.
Risk levels
How websites are rated
For websites, the rating you see in the MFA data feeds and in the MRT depends on how many signals are high:
- Medium: one of the four website signals is high.
- High: more than one of the four website signals is high, or the site serves an extremely high volume of ads (for example, extremely high ad density).
“High” and “extremely high” are measured against Pixalate’s global traffic pool.
What to do with each level
Pixalate recommends transacting only with properties rated Low MFA risk. Treat both High and Medium as disqualifying when you source inventory. The difference between them reflects how clear the signal is, not a difference in what action to take.
Definitions
MFA (made for advertising)
Ad-enabled websites, CTV apps and mobile apps that are intentionally designed to generate arbitraged traffic. These properties generally have high ad density to maximize revenue, with little regard for user value.
MFA risk
Pixalate’s assessment of the risk (high, medium or low) that a website, mobile app or CTV app is MFA. See Risk levels.
Ad refresh rate
The number of ad impressions observed per device, per minute.
“Ad refresh rate” is Pixalate’s own term here. It doesn’t mean the rate at which an individual ad placement refreshes. To bring it down, look at the total volume of ads displayed and how quickly they’re triggered. That can be a combination of the number of ad placements on the page and when the ads fire. For example, ads that fire before the user has a chance to see them push the rate up.
Ad density
The number of ad placements observed per device, per minute. Websites only.
Paid traffic rate
The volume of traffic sourced through paid (ad) sources. Websites only.
Social traffic rate
The volume of traffic sourced through social media. Websites only.
Age
The age of the app in its app store. Mobile and CTV apps only.
IVT
The app’s invalid traffic (IVT) percentage across regions. Mobile and CTV apps only. For Pixalate’s IVT types, see the IVT taxonomy.
Popularity score
Pixalate’s popularity score for the app. Mobile and CTV apps only.
Reviews
The number of reviews the app has. Mobile and CTV apps only.
Where to use MFA risk
- MFA data feeds: lists of medium- and high-risk websites, mobile apps and CTV apps for pre-bid blocking. The feeds are updated daily.
- Media Ratings Terminal: MFA risk in Discovery and on a website’s Insights page.
- Enrichment API: the
mfawidget, for looking up one property at a time. - Analytics dashboard: your exposure to MFA traffic.
For an overview of MFA and why it matters, see Made for Advertising (MFA).
Questions? Email support@pixalate.com.